Sixty-two adults finished a home sleep experiment with a headband they put on themselves. The team had enough usable data from 49 of them to compare timed sounds with a sham condition. Looking at which nights survived that second step makes the result easier to interpret.

THE SHORT READ
  • During usable recordings, timed sounds produced a measurable electrical response and increased power in the slow-oscillation and spindle frequency bands.
  • Of 490 experimental nights from the 49 participants retained in the analysis, 319 passed the recording criteria. That is 65.1%.
  • The study detected no change in the overall distribution of sleep stages. Whether people feel more rested during the day remains an open question.
THE STUDY AT A GLANCEHome acoustic stimulation with wearable EEG · Černý et al.
Publication date
22 November 2026
Protocol completed
62 adults with self-reported symptoms
Usable nights
159 stimulation and 160 sham
Main measurement
Electrical brain activity during sleep

Original publication: 22 Nov 2026 · The date above refers to this brief.

THE NUMBERS, IN CONTEXT

Which night slots reached the EEG comparison?

night slots

Scheduled for all 62 completers620
From the 49 included participants490
Retained for the comparison319
0650
These bars are nested. Each completer had ten experimental nights scheduled, excluding habituation.
THE EXPERIMENT, EXPLAINED

From a person to an EEG result

  1. Complete the schedule

    Each person had five nights in each of the two conditions.

  2. Apply the recording criteria

    Keep qualifying nights for people with usable data in both conditions.

  3. Compare the signals

    Compare electrical responses in stimulation and sham recordings.

Press-News diagram of the reported recording and inclusion steps.

A night in the study had to pass two checks

Adults reporting trouble staying asleep wore the device at home. After two nights to get used to it, each person had five experimental nights under each condition. The order was counterbalanced. Entry into the final comparison required at least one usable night in each condition. A recording could fall short because of poor signal or too few qualifying events during sleep. Thirteen of the 62 people who completed the protocol fell outside the final analysis.

Source 1 ↗

Two percentages describe different parts of the experiment

The 49 included participants supplied 490 experimental nights. The team kept 319, which gives the reported 65.1%. For someone planning a similar study, the original schedule also matters. The 62 completers had 620 experimental night slots between them. The same 319 nights represent 51.5% of those slots. This second calculation describes the overall yield from the schedule. It includes the slots of people who did not qualify for the analysis. It cannot assign a technical fault to every unused slot.

Source 1 ↗

A signal response and a better morning are separate outcomes

Compared with sham, timed sounds changed the EEG pattern around the event and increased power in two frequency ranges associated with sleep. The authors detected no shift in the overall sleep-stage pattern. A recording can show that a sound influenced electrical activity without telling us how rested someone felt the next morning. A follow-up study could pair the overnight signal with a prespecified measure of daytime functioning and track people whose recordings are difficult to collect.

Source 1 ↗

What changed when the experiment moved home

In an earlier laboratory experiment, some of the same researchers recorded healthy sleepers with the headband and with simultaneous polysomnography during one night. The two systems both picked up a response. The home study tackles another problem, whether people can supply usable data repeatedly in their own rooms. The authors disclose employment or financial interests in Deep Sleep Technologies and an inventor’s interest in the method. Independent replication could test both measurement quality and patient-reported outcomes.

Source 1 ↗Source 2 ↗
TRACE THE EVIDENCE

Trace the denominator from a completed schedule to an analysed night

Publisher abstract and article sections covering recording retention, Figure 3, the EEG results and conflicts of interest. No participant-level recordings were examined.

01How many nights informed the comparison?

What was observed
159 stimulation nights and 160 sham nights passed the criteria in the included participants.
Where the conclusion stops
Several nights can come from one person.

Source 1 · Results on recording retention and Figure 3

02What changed during stimulation?

What was observed
The authors report an EEG response and higher power in the slow-oscillation and spindle bands.
Where the conclusion stops
The overall sleep-stage pattern showed no measurable change. Daytime benefit remains a separate question.

Source 1 · Abstract and electrophysiological results

Numbers you can inspect

MeasureValue & unitOrigin & method
Protocol completers62 peopleReported Completed the home protocol.
Source 1 · Recording retention, Figure 3
People included49 peopleReported Usable data in both conditions.
Source 1 · Recording retention, Figure 3
Completers included79 percentCalculated 49 divided by 62 times 100, rounded to one decimal.
Source 1 · Recording retention, Figure 3
Planned night slots in all completers620 night slotsCalculated 62 completers times ten experimental nights. Excludes habituation.
Source 1 · Study design and Figure 3
Nights in the retained participants490 nightsReported 49 participants with ten experimental nights each.
Source 1 · Recording retention, Figure 3
Usable nights319 nightsReported All retained nights across both conditions.
Source 1 · Recording retention, Figure 3
Usable stimulation nights159 nightsReported Retained in the stimulation condition.
Source 1 · Recording retention, Figure 3
Usable sham nights160 nightsReported Retained in the sham condition.
Source 1 · Recording retention, Figure 3
Yield in retained participants65.1 percentReported Authors’ reported percentage using 490 nights as the denominator.
Source 1 · Recording retention, Figure 3
Yield across the schedule of all completers51.5 percentCalculated 319 divided by 620 planned experimental slots times 100, rounded to one decimal.
Source 1 · Study design and Figure 3

Compare the actual experiments

These studies answer different questions. Read the unit and endpoint before comparing results.

StudyUnit & settingReadoutInterpretation boundary
Single-night laboratory comparison

Source 2 · Published abstract, Methods and Results

Healthy participants with headband EEG and concurrent polysomnographySignal response on two recording systemsThe earlier laboratory comparison checks measurement agreement under supervision.
Repeated home recordings

Source 1 · Home study abstract and Figure 3

Adults with self-reported sleep-maintenance symptomsSignals recorded on eligible nightsHow people feel or function requires a separate measure.
Download evidence table (CSV)

The export includes claims, available numbers, methods and source locations. It contains our reading notes and published summaries; it is not raw participant data or an independent reanalysis.

Evidence update · 22 Nov 2026
First publication. Any calculations in this brief use reported figures. The reading notes identify possible follow-up questions. AI source check; no human editorial or clinical review.

CONNECT THE EVIDENCE

Three ways to read the same experiment

UnitCalculationInterpretation
People49 of 62 completers, or 79.0%Participants represented in the EEG comparison.
Nights in the retained group319 of 490 experimental nights, or 65.1%Recording yield among the included participants.
Scheduled nights in all completers319 of 620 night slots, or 51.5%Yield across all ten-night schedules. Habituation nights are excluded.

Press-News calculated the participant and all-schedule percentages. Figure 3 reports the night counts and the retained-group percentage.

READER QUESTIONS

Your questions, answered

Were the 319 nights from 319 different people?

They came from 49 people. Each could contribute several nights in each condition.

Did the sounds make people spend longer in deep sleep?

The researchers detected changes in the electrical signal. They reported no measurable change in overall sleep architecture.

Why calculate 51.5% when the paper reports 65.1%?

Both use 319 retained nights. The paper’s percentage uses 490 nights from included people. Our figure uses all 620 experimental slots scheduled for the 62 completers.

What did the earlier laboratory study check?

The researchers compared signals from the headband with simultaneous laboratory polysomnography during one night in healthy participants.

LIMITATIONS

Limits of this interpretation

  • The sample was defined by self-reported difficulty maintaining sleep.
  • Nights that failed recording criteria and people without qualifying data in both conditions are underrepresented in the EEG findings.
  • Sleep-stage composition and daytime functioning are different outcomes. The authors disclose financial and patent interests.
SOURCE NOTES

Sources & transparency

  1. Černý et al. (2026) — Unsupervised Acoustic Stimulation at Home in Users with Self-Reported Maintenance Insomnia Symptoms

    Publisher abstract and article sections covering recording retention, Figure 3, the EEG results and conflicts of interest. No participant-level recordings were examined. · Accessed 22 Nov 2026

    DOI: 10.3390/s26237381
  2. Jongejan, van der Heijden and Talamini (2026) — Closed-Loop Acoustic Stimulation of Slow Oscillations Using a Wireless EEG Headband

    Published abstract. Checked the single-night laboratory design in healthy participants, the simultaneous polysomnography reference and the reported signal response. · Accessed 22 Nov 2026

    DOI: 10.3390/s26206528

Prepared and source-checked with AI for Press-news Team. The source notes identify the material available during preparation. The article received no human editorial or clinical review. The source list supports the reported results and the calculations shown.

Source check: AI source check — primary research and stated source limits

Clinical review: Not applicable to this educational guide

Suggest a correction